n8n-io
Development Feed
Development Feed
The data redaction module now initializes automatically on startup, removing the need to set an environment variable to unlock this privacy feature.
n8n workflows can now report AI token consumption per node to telemetry, giving visibility into how many tokens each AI interaction uses.
Workflow data redaction settings now require an enterprise license. Without it, the options are visible but disabled with an upgrade prompt.
The token exchange module now validates and loads trusted public keys during startup, rejecting misconfigured keys before they reach production.
A bug that caused the wrong API key to be passed to Daytona sandbox proxies has been resolved — proxy mode now fetches a fresh auth token instead of reusing a stale static key.
Webhook-only triggers like Chat Trigger in Instance AI can now be tested directly — previously, clicking "Test" silently did nothing instead of running the trigger or returning a meaningful error.
Large AI tool results that crash browsers when opened can now be downloaded in full — the interface truncates display at 100K characters to keep things snappy, with a one-click download for the complete data.
Automated end-to-end tests now run nightly against the VM-based expression engine, catching expression-related bugs before they reach production.
CI scripts for version bumping, changelog generation, and GitHub releases have been modernized with better transitive dependency handling, improved commit filtering, and new automation for release creation and promotion.
Merge queue events no longer trigger end-to-end test suites, cutting CI noise and compute waste when rebasing in GitHub's merge queue.